Evals for Every Language - Latvian (lv): leaderboard
Average score across evaluation tasks in Latvian (lv), from Evals for Every Language.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 77.15 | #49 |
| 2 | Gemini 3.1 Pro (Preview) | 76.21 | #58 |
| 3 | GPT-5.1 | 74.81 | #135 |
| 4 | GPT-5 | 74.25 | #98 |
| 5 | GPT-5.2 | 73.71 | #110 |
| 6 | DeepSeek V3.1 | 73.18 | #264 |
| 7 | GPT-5.3 | 73.02 | #103 |
| 8 | Claude Sonnet 4.5 | 72.94 | #142 |
| 9 | Claude Opus 4.8 | 72.21 | #37 |
| 10 | GPT-5.5 | 71.94 | #31 |
| 11 | Claude Sonnet 4 | 71.05 | #199 |
| 12 | Claude Sonnet 4.6 | 70.98 | #90 |
| 13 | Nova Pro (v1) | 70.67 | #497 |
| 14 | GPT-4o | 70.66 | #343 |
| 15 | Mistral Medium 3.1 | 70.33 | #408 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-lv · How It Works · Data refreshed daily, snapshot 2026-10-20.